In the past, it was common for enterprise IT professionals to build, own, and operate their own data center to accommodate growing business needs. Although this provided full access to the data center and allowed companies to avoid leasing problems and control their own temperature and cooling, many underestimated the costs and risks involved in building and operating a data center. Others found that they didn’t have the specific skillsets, expertise, and operational mindset required to avoid costly data center outages.

Every year, as more and more companies rely on data centers to keep their business running, the consequences of data center outages increase exponentially. According to Uptime Institute’s Annual Industry Survey, nearly half of Enterprise IT organizations experienced an outage in their own data centers over the past year. For many businesses, the benefits of owning and operating a data center isn’t worth the risk and expense.



The cost to design, build, and operate a data center is astronomical. If no one on your staff is a data center design expert, you will need to hire consultants to find a geographic location that mitigates against risk from natural and man-made disasters, is close to nearby emergency routes and roads, has multiple network carrier options, and access to power sources.

Once you find the most ideal location, you will need to hire specialized engineers, designers, contractors, and architects to design and construct a data center that meets the business requirements for reliability and resiliency. Simply stated, you must ensure the data center is designed and constructed to provide the critical systems uptime needed to keep the servers powered, connected to the network, at the optimal temperature, and mitigated against every conceivable risk.

When a data center’s design doesn’t take every detail into account, disaster can strike. An article on Computerworld UK highlighted an incident in 2011 when Microsoft and Amazon’s European cloud services went down for an entire weekend after its data centers in Dublin were hit by lightning. Both the main power and the backup generator failed, taking businesses on AWS EC2 platform and Microsoft’s Business Productivity Online Suite offline.



According to a study conducted by Forrester, the estimated construction cost of a data center starts at approximately $200/ sq ft, although it frequently costs much more. Additionally, the installation of fiber can cost upwards of $10,000 per mile to reach your site location. While the up-front costs for building a data center’s infrastructure can vary, choosing lower quality options to save money will usually cost more in the long run because the components will have to be replaced more frequently. Replacing equipment more frequently disrupts the systems, which makes the data center more vulnerable to an interruption in service.

A article details two outages in Amazon Web Service’s Eastern U.S. region that affected cloud services as well as sites like Heroku, Github, and CMSWire. Amazon’s repeated failures in the region are said to be related to their data centers’ aging infrastructure.  



The most expensive, difficult, and intensive part of owning and operating a data center begins once it has been built. Powering the data center accounts for 70% to 80% of the overall cost of running the facility. Although the industrial utility power rate varies greatly region to region, in March 2016, the U.S. Energy Information Administration calculated the national average to be 6.47 cents/kWh. The average daily consumption of a 5,000-sq-ft data center is 27,048 kWh, which is around $1,750.01/day and $54,250.17/month.

Staffing your data center with the right employees is another hurdle. The best way to mitigate against downtime caused by human error is to hire staff members with the specific skillsets, expertise, and discipline needed to maintain a data center’s critical systems uptime. Once you’ve hired an operations team, you’ll need to create and enact a training program to instill a strategic and operational mindset in each employee.

An outage at a NASA data center speaks to the value of a comprehensive training program. NASA’s website reports that on August 26, 2011, a fire suppression sprinkler head leaked at the NASA/Caltech Jet Propulsion Laboratory (JPL), eventually triggering the full water release when the maintenance crew responded. The affected offices share the fire alarm system with an adjacent data center that, among other activities, supported Mars Science Laboratory (MSL) Assembly, Test, and Launch Operations (ATLO). The fire alarm system was programmed to shut down power to the data center unless the shutdown was manually countermanded within 10 minutes. All of this occurred during non-office hours and the security guard, not knowing any better, arrived after the 10-minute period had elapsed and the data center suffered a power outage.  

Owning and operating a data center is expensive, resource intensive, and comes with many pitfalls — the biggest being a costly outage. It requires focus, specific skillsets, and an operational mindset that few truly understand. The endeavor goes far beyond a construction project aimed at gaining a little more control over your organizations IT environment.

Alternatively, there are many data center service providers across the country that have weighed the risks, taken the total cost of ownership into account, and dedicated themselves fully to delivering continuous uptime to their clients. Colocation customers have access to a more predictable expenditure model, a flexible environment that can scale up or down quickly, and the data center’s expert staff.

For many companies, the benefits of leasing data center space outweigh the many drawbacks involved with building a data center of their own.